Presentation Title

Woolsey Fire GIS Land Classification Using Aerial Imagery

Faculty Mentor

Omar Mora, Steve Steinberg

Start Date

23-11-2019 1:00 PM

End Date

23-11-2019 1:15 PM

Location

Markstein 102

Session

oral 3

Type of Presentation

Oral Talk

Subject Area

engineering_computer_science

Abstract

Aerial imagery was used in identifying and classifying changes in the areas affected by the Woolsey Fire that took place in November 2018. The aim of this work was to determine the percentage of damage that affected the structures and vegetation after the fire occurred in order to determine the possibility of structures repair and regrowth of vegetation. With that imagery and ArcMap software, different types of land cover such as residential and commercial areas, infrastructure (roads, utility stations, etc.), and vegetation were analyzed using both unsupervised and supervised classification processes. Because the software did not provide an accurate classification of ground cover using its default bands (red, blue, green), different variations of settings were tested to find an optimal configuration to accurately classify the ground cover. After a desired level of accuracy had been achieved, these processes were used to determine the area of each land type in the region, which affected areas in both Los Angeles and Ventura counties. To provide an evaluation of the damage caused by the Woolsey Fire, imagery from before and after the fire were compared. After collecting, evaluating, and editing training samples, the results show that using supervised classification allowed better accuracy in the identification of the structures and vegetation affected by the fire.

This document is currently not available here.

Share

COinS
 
Nov 23rd, 1:00 PM Nov 23rd, 1:15 PM

Woolsey Fire GIS Land Classification Using Aerial Imagery

Markstein 102

Aerial imagery was used in identifying and classifying changes in the areas affected by the Woolsey Fire that took place in November 2018. The aim of this work was to determine the percentage of damage that affected the structures and vegetation after the fire occurred in order to determine the possibility of structures repair and regrowth of vegetation. With that imagery and ArcMap software, different types of land cover such as residential and commercial areas, infrastructure (roads, utility stations, etc.), and vegetation were analyzed using both unsupervised and supervised classification processes. Because the software did not provide an accurate classification of ground cover using its default bands (red, blue, green), different variations of settings were tested to find an optimal configuration to accurately classify the ground cover. After a desired level of accuracy had been achieved, these processes were used to determine the area of each land type in the region, which affected areas in both Los Angeles and Ventura counties. To provide an evaluation of the damage caused by the Woolsey Fire, imagery from before and after the fire were compared. After collecting, evaluating, and editing training samples, the results show that using supervised classification allowed better accuracy in the identification of the structures and vegetation affected by the fire.